PHYSIOLOGICAL CO:\,CEPTS OF co:\'scrOUSi\ESSn,

R. K. ANAND

(From Iht Dtfmrlm.nl ~j PhyJi%,V. Ail·f,ldil1 IIIJlilult oj "1,dieD/ Sr;l1IuJ,N,w Dtihi.)

Rm;;;td en Jul.), 10, 1960

In the words of Russell Brain (1958), "there are no more complex prob­lems in neurology than those which arise in connetion with consciousness, andfew topics more plentifully breed confusion". Evcn to define the m<-aningof the word consciousness is not easy. \VilliamJames (1892) said that every­one knows what consciousness is until he tries to define it. It is, therefore,no use here to try to define it. The definition given by StanJe}' Cobb (1948)that consciousness is "awareness of environment and of self" may be sufficientto convey the meaning of this word.

Various other wOros are also used by physiologists and other workers todescribe the state of nervous system, which is synonymous with cunsciousness.These would be the states of wakefulness, altention, alertness or arousal of thebrain. It witl be observed later on that the activity of the brain and themechanism producing it in all these states is similar. Opposite to this wouldbe the states of sleep or unconscious~ess. The brain mechanisms producingthese states are again similar. The unconsciousness resulting in anaesthesiaor in coma agaill are related phenomena, with the only difference that it isp::mibJe to wake up a sleepin~ person, while it is not so in the other condi­tions. The reason for this will be obvious when the mechanism producingsleep and wakefulness, or unconsciousness and consciousness are understood.

Two recent reviews have appeared on this subject, one by Magoun(1958-"The Waking Brain.") and lhe othor by Feldberg (1959-"A physio·logical approach to the problem of general anaesthesia and of loss ofconsciousness") which have very much clarified our concepts. These twO

reviews have mainly fOlllled the basis of the present review.

I .- Cerebral cortex and consciousness

Por a long time it was supposed that consciousness was exclusively a func­tion of the cerebral cortex. The unconsciousness produced as a result of headinjury or concussion, was attributed to cortical damage. Neurologists believedthat the loss of consciousness from cerebral haemorrhage was caused by gene·ral cortical anaemia.

In animal experiments, on the other hand, it lVas observed as early asthe end of last century (Goltz, 189::!; Schafer, 1900), that removal of cerebralhemispheres does not cause loss ofconsciousness, and these animals can sleep

156 CONCEPTS OF CONSCIOUSNESS

as well as awake. Cairns (1952) showed similar responses in human anence­phalic and hydrocephalic monsters,

Jefferson (1938) pointed out that the neurologists WCfe wrong in assum­ing that the cerebral cortex was the seat of cotlsciousne~s. He and his collea­gues (1950,1958) published maps on the basis of clillical cases, showing thatdamage to the upper brain stem and diencephalon leads to depression ofcons­cioumess, hypersomDia or coma.

The real knowledge regarding the involvement of upper brain stem, andits relationship with the cerebral cortex, in the production of the states of"consciousness", on theOlher hand, has emerged as a result of a large numberof experimental studies conductt:d by nt:urophysiologists. I will now pass onto dt:scribt: somt: of tht:m which havt: acted as rt:allandmarks in the t:voluriollof our knowledge regarding this mechanism.

II :-Classic contributions to the problelD.

(1) EEG pallerns in relation to consciQI/Sntss ond sitep.

Shortly after discovery of the electroencephalogram by Berger (1929),came his obsef\'ation that in sleep its patterns tended to be composed of large,slow, wavelike fluctuations of potential (Gamma \Vaves), while a flattenedrecord was characteristic of wakefulness. Thus the EEG patlt'rn of sleep con·sists ofhigh-voitage, slow-waves of synChronised discharges, while duringwakefulness or alertness the record consists of low voltage, fast, desynchro­nised activity (Fig I). This latter type of desynchronised record is nowaccepted to be synonymous with the 'arousal' of the animal. The high­voltage slow-wave record is obtained during sleep, as wt:ll as unconsciousnessproduct:d as a result of anything.

~_rv""'<JJ'"",wrv..rV'vo"""----~-----­M

,.

tHAND-CLAP CAT OPENS EYES AND RAISES lIE... D

SLOW WAVES OUT Ii MIN. SPINDLES OUT 2! MIN.

Fig. I. EEG of a cat when lllIleep and when woken up. (Fr,)m Lindsley et at, 1950).

Rheinberger and Jasper (1937) studit:d silllultaneons EEG and behaviourof cats and observt:d that tht: low-voltage fast pattern of EEG occurs along­with attt:ntivt: wakefulness and is tht:rdore termed EEG 'activatiol,' or 'arou­sal'. This can be evoked equivalently by stimulations of several afferentmodalities (Fig. 2). However elicited, this is distributed generally over thewhole hemisphere and not restricted only to the brain area which reCeives

B. K. "NAND 157

the afferent sensation. Also, once aroused, this 'arousal or activation patlern'lends to persist for a long period, as long as lhe behavioural 'arousal' (atten­t;\'e wakefulness) last~. These very important observations indicated theimportance of afferent stimulation in initiating both EEG and behaviouralwakefulness.

J. SOUND

A 2 I3 [

4[

B 2131

4[

Fig. 2.

~. "'f'l: ..:/\ j~.1\ '#', .:~

J. SOUND

EEC from S/:\'cral regions of the cerebral hcmisphere of an inallentil,e (drowsy) cat,showing generalised activation pattern evoked by auditory stimulation. (Froll'Rheinbcrgcr and Jasper, 1937).

These EEG studies then formed the basis or most of the important stu­dies which have been carried out to elucidate the mechanism of productionof consciousness and sleep. In some experiments wakefulness has been studiedas a behaviour change and this is termed as 'behavioural arousal or wakeful­ness'. But such studies can be conducted only in unanaesthetised animals.Therefore majority of the studies have utilised 'EEG arousal' and EEG sleep'to interpret these states in lightly anaesthetised animals, in whom behaviouralarousal can not be pro{(uced. High.voltage slow synchronised waves areinterpreted as synonymous of the state of sleep or unconsciousness, whilelow-voltage fast dcsynchronised waves are interpreted as synonymous of thestate of arousal, wakefulness, consciousness, alertness etc. Thus the pioneerstudies with EEG cannot be underestimated in terms of their importance forthese studies.

(2) Brain Jtem mtthaniJTrU in relation 10 couJciollJneJJ and Jlttp.a) Studies by Bremer:-

Bremer (1935, 1036, 1937) made two typ~s of cats with transverse cutsthrough the brain stem preserving the blood supply to the cortex and studiedtheir behaviour and EEG. He observed that when sections are made between

15B CONCEPTS OF CONSCIOUS!'>!':S!

the superior and inferior colliculi (8 of Fig. 3), the head of such a cat behavt:Sas in constant sleep and the EEG record shows typical high-voltave slow·I\ave slet-p-palter'o. This preparation was callcod "ClTI'fOtl /sole". On theolhu hand, if the seerians are made below the medulla oblongata (1\ ofFig. 3) leaving the brain in conlact with the brain stem, such a prt:paralionexhilJits mainly periods of wakefulness which alternate with periodsof sleep. During wakefulness the EEG shows low-voltage fast desyn.chronised waves, \... hile during sleep the EEG record is typical of sleep. Thispreparation was named "Elluphale [sofe", Bremer. therdore, concluded that

D

- B

•Fig. 3. EEC or cal sho-, jog waki"s record (A) of t~uphale isol, following section at bull>o!­

pinal, juncture (D_A) and !l~ping record (II) of rtrt'I/1U isol, following mid·collicular~ccli(>n (D-B). The sleeping appearance of the latter il seen in (e). (From Bremer1937).

sleep was the constquenee of deafferentalion (CUlling out of sensations) of thecerebral hemisphere. Later studies mostly wppontd the general featurts ofBremer's conclusions, with this dilTe~nce that it W,lS shown that the sleeppattern is nOI due to cutting out of the direct sensory routes to cerebral cor­tex, but due to cUltin~ out of the projections from the central reticular forma­tion of the brain stem.

•

•

B. K. ANA 0 159

(b) ludies b)I Mooou//'s group:-

"e owe our pre ent concept about the mechani m of consciousnemainly due to ex len ive tudie which ha e been carried out by Iagoun(1 -8) and his group. 10ruzzi and Maaoun (1949) found that direct timu­lation of the central reticular ~ rmation of the brain stem reproduced all thee1ectrocortical feature obsened in tbe EG arou al reaction associated \ ithwakefulne. uch tudi s have been often r peated ince then and haveh Iped to elucidate a number of fact reoalding the mechanism of wakeful.ne. egundo et at. (185-), \ ith the help of chronically implanted electro­de in the r ticular formation, bowed that reticular stimulation evokes thesame pallern of EG arousal as is induced by peripheral eo ory stimulationand this coincides with behavioural awakening (Fig. 4).

CING.

R.F.

MONKEY QUIET, EYES

CLOSEDA D ON-R'S­

PONSIVE

MONKEY Movr 'c, EYE OPEN AND

RESPONSIV,

-I NOISE

CING.

'A.F.

-D STlM. R. F.

OUR G

12 ", r MSEC., 200 I SEC.)

-ig. 4. EEG and motion picture frames from a monk y who was aroused with peripheralafferent stimulation (I. oi e) and by stimulation of reticular formation. Thearousal in both cases is identical. (From egundo et al., 191 ). (II TM R.F.)

hese tudies were further extended by observing the consequences ofexperimental destruction involving the brain stem and diencephalic regions(Lind leyet al., 1950, French et al. 1952). ni'mals with large Ie ion of thecentral cephalic brain tern remain a though deeply asleep. Their E Grecords show sleep or coma waves. either behaviour, nor EEG arousal canbe produced in them by any peripheral sensory stimulation. ven if the long

16') CO",CEPTS OF CONSCIOUS~ESS

sensory and motor paths to the cerebral cortex passing throup:h the brain st('mare spared and only the cenlral core of reticular formation is destroyed, suchanimals show no signs of awareness of their environment, and cannot showeither behavioural or EEG arousal. On the other hand. if ollly the lon~

sensory paths to the cortex rae destroyed and the central core ofretic.ular for.mlltion is inlact, the animals are capable of both behavioural as well as EEGarousal. It was seen in our laboratory (!\nand, 1955) that small destructivelesions in the mammilary region of the hypothalamus produce typical coma­tose picture and it was therefore concluded that the mammillary region mayform a very important component of this reticular activating system.

Thesc studics dcmonstrated that the afferent influx tral15ported cCOlrallythrough thc brain stcrn is implicated in the arousal mechanism. Thc comaof ((rutau iSQle animals, therefore, depcnds upon cxclusion from higher corticalstructures of 'activating' stimuli conducted selectively through this medianzone rather than upon the blockade of primary sensory signals transported tothe cortex through the lateral brain stem (French, 1960). The area impli­c,lled in conveying such '·arousing" information to the brain is occupiedprincipally by the reticular formation and related thalamic nuclei; hcncethese struclUres have come to be known as "reticular activating system"or RAS.

III. Relation of Reticular Activiting Systetn to Consciousness.

(I) Anatomical considualiol1,f.

The reticular formation proper begins in the medulla a little above thedecussation of pyramids, and extends through lhe whole of brain stem to thebase of brain. It is centrally located, being surrounded by a ~he" of neuraltissue consisting of long fibre tracts and nuclei. It contains groups of neuronswith interspersed fibres. Allen (1932) observed that in illl development, thereticular formation surrounds the sensory nuclei of thalamus, and such struc_tures as the red nucleus, substantia nigra, or mid brain lJuclei and parts orhypothalamus should be considered probably as specialised derivative.~ of it.

The 'reticular activating system' includes the cephalic portions ofreticu­lar formation, as well parts of thalamus, subthalamus, epithalamus andhypothalamus. This unit constitutes the mechanism for activation of thecerebral cortex.

(2) Ascending conntdion.1.

This diffuse system ia the middle of braIn stem (RAS) receives sensorysignals from collaterals of all kinds of ascending fibres. As the classicalafferent paths ascend towards the cortex collaterals pass widely from theseinto the central region. Thcse collaterals are received from the somato_sensory fibres in the medial lemnisci, from the other cranial nerves, and frOmvisceral afferent fibres. The reticular formation in turn sends impulses

B. K. ANA:>;'D 161

to practically the whole cortex. These impulses desynchronise the high­voltage slow. wave activity of the cortex thus activating it. This arrangementis illustrated diagramatically in Figs. 5 <Ind 6.

Fig. 5. Later.. l view of monkey's brain showing c1a",ical somatic afferent pathway projected

to the se''IIory cortex. and lhe reticular activating syucm in lhe cemral brain sternreceivin~ collaterals from the classical afferent pathway and lhcn projecting diIT"se]\"to the cortex. (hom Magoun, 1954).

(3) Muhm.ism of aronJal.

It is thus evident that arousal of the brain is a function or the R/\S.\Vhen the Rt\S is not activated, the rhythm of the brain is s~'nchronised slowhigh. voltage. During this stage any afferents which may go up to the cortex.are not well received as the cortex has not been alerted to receive them.-On lhe other hand, when the stimulation of peripheral afTerents acti~·ates theRAS through the collaterals going to it, the activated RAS arouses the brain

'through its diffuse projections. During desynchronised fast low-voltageactivity, the cortex has been altered and is ready to receive and interpret anyperipheral afferents which will project to the cortex, Thus it will be seenthat the stimulation of the peripheral afferents themst'lves will arouse thebrain through collaterals going to RAS and thus alert the brain for their own..eception and interpretation. It may be noted here that the arousal mecha­.nism is mediated through multiple ,rdays (synapses) in the RAS and th-us

162 COSCEPTS OF C(}~SCIOUS~ESS

exposed to the effect of certain factors (anaesthetics for exam pie) at thislevel.

(4) TYPe! ofafJarenl stimuli capable of arousal.

The RAS is capable, probably of some spontaneous or autochthonousdischarge, although the bulk of its tonic potency is derived from in severalinputs. However, some receptor systems exert a more powerful excitatoryinfluence upon the RAS than do others. Stimulation of the visual nerves hasbeen found to be least effective (Arduini tt al, 1933), while somatic sensoryeltcitation is most poten! in this regard (French eJ al, 1952). It has beenshown that even in encephale isoie Ihe wakefuln::5s depends upon sensory inputs10 the Rt\S from the trigeminal nerves.

•

[3

•THALA~llC

NUCLEI

R~T1CULAR

ACTIVATINC

svnEM

Fig. 6. Diagram or brain showing classical somatic afferent l"l.Ihway projecling to :rensorycorlex (!IOlid line), aod rcticular activati:1g system (black) reechoing colltaeral~ andWilh diffuse cortical projections. (From French, Vcncano and l\lagoun 1953).

(5) Parl played by cerebral cortu in arousal,

Although arousal is produced through the RAS, the activity generatedthere must be exerted upon the cortex, in order ror the alert state to be dis­played. The cortical regions contribute to the process of arou~al hy func­tioning in the maintenance of prolongation of sensory-induced awakening(French 1952). Decorticate animals and man exhibit only transient brierperiods of apparent wakefulness, but during such temporary arousal, alertnenor appropriate reaction to environment i, impossible. It appearss therefore.

B. K. ANAND l6~

that crude arousal is possible without cortical contribution, hut the intactcortex is ~ssential for prolonged sustained alert wakefulness characteristic ofthe normal adult subject.

(6) Phylogenetic de/)eloplJI~nl ofCOf/u;ousness.

If one studies the phplogenetic development and evolution, one comesacross the same phenomena which have been elucidated above. In the earlyevolution, with the development of nervous system analagous to spinal cordand brain stem, the animal develops a general "awareness" of its environ­ment. On top of these developed the older parIS of brain (Rhinencephalon-Limbic System), which now are considered very important for regulationof visceral acliviti~s. This also regulates the affective behaviour and so herewe have the "feeling" in response to the enviromental stimuli. Then develofedIhe neocortex, and Wilh this came the "knowledge" of the environmentalstimuli. This is what we understand by the term consciousness. It will beapparent from this also that the basis of this "knowledge" or consciousnesscould be the activity at brain stem level, producing general awareness.

(7) Centrtllctphalic Syslem of Ptnfitld.

By definition this refers to neurone ~ystems which connect both hemi­spheres and which are located in the central core of the upper brain stem.According to Penfield (1952) this centre in the upper brain stem is the seat ofall consciousness, and through which all activities of the two cortices are in­tegrated. Although differing in certain important respeclS, the basic conceptof this that the ~eat of consciousness is in Ihe brain stem, fits in with thehypothesis elaborated above.

(8) Relalionship 10 pain condl/clion.Collins el al (1954) a nd Livingston (1954) have shown that stimulation of

peripheral afferents subserving pain, evoke potentials in the mid brain teg­mentum. These studies suggest thaI, from the bulbar level forward, theascending pain pathways may be made up very largely of relays through theRAS. They would thus be sensitive to central anaesthetics here (see below).It also seems appropriate that a central neural mechanism, concerned withalerting and arousal, should be supplied abundantly with connections fromtht: afferent pathway for pain.

IV. Wakefulness and Sleep.

(1) Wakifulnm.

Dile can now summaries wakefulness as a condition of the brain whichhas been aroused or alerted by the RAS, which in turn has been activatedby projection of afferent stimuli-mostly somatic into it. Such an alertedbrain will receive and interpret the afferent stimuli direct by projecting to

164 CONCEPTS OF CO;";SCIOUS:"F.SS

the cortex and in such a state we will, therefore, be aware of our environ­ments. \Vhen this arousal of the cortex is lacking, the afferent stimuli pro­jecting directly (0 the cortex will not be aware of our environments i. e. wewill be unconscious.

(2) Siup.

This condition is produced when the cortex is not aroused by the RAS.II will, therefore, occur when no peripheral afferents arc activating the R.\S·In this condition the mechanism of RAS is still intacl, only it is not beingactivated. If any peripheral Slimulus is applied and the afferents arc strongenough to activate the RAS, the condition of sleep will change into wake­[ulnc,s.

!'ig. i. Stimulation of anteromedial amygdala through implanted ekclrodcs produce! skcpin cat. (From Anand and Dua, 19:16) .

Previously many other hypotheses had been put forward to explain themechanism of sleep, but it is no use referring to these here. There are certainstimulation studies, however, which at first sight do not seem to conform withthis general concept regarding sleeo and wakefulness. Sleep according toIhis is due to blockage of impulses going up to cortex from RAS. I-less (1941-),on the other hand, showed that cats, in whom medial diencephalic regionswere stimulated through implanted electrodes, went into normal sleep durin£:,Ihe period of stimulation and so he had suggested a sleep centre in thediencephalon. It has now been argued that this sleep is produced by slimu·

-

fl. K. "NANC' 166

lation at slow rate! of medial thalamic nuclei, which produces delta sleepwav~s in the corlex. In OUI' laboratory also it has been observed that stimu­lalion of amygdaloid nuclei sometimes induces sleep in cats (Anand and Dun1956, Fig. 7). It is suggested lhat there may be other nervous mechanisms;:,Iso which could block the RAS projection to -:ortex and thu~ produce sleep.

3. Ul/col/sciousntJJ find Coma.

This would be a sleep like condition produced again by blockage ofimpulses projecting from RAS to cortex. This would be produced by anycondition which dama.qes the reticular formation in the brain stem ordiencephalon. This differs from sleep ill the essential feature that such anindividual cannot be aroused with the help of peripheral sensory stimula­tion. This is obvious, as in this case the RAS has been damaged and socannOI arouse or alert the cortex; which, therefore, stays in a continuousst,Lte of sleep and unconsciousness.

4. Val/Mgt to Ctrebral Cor/tx.

t\s has been explained above, in such a case although the RAS which isresponsible for wakefulness is intact, the cortex which displays the alert staleis not functioning. Such individuals can exhibit only transient brief periodsof apparent wakefulness, but alertness or appropriate reaction to environmentis not possible.

V. Effect of Anaesthesia.

The loss of wakefulness in central anaesthesia has been attributed to adepression of activity within the RAS. Under light anaesthesia it has beenconvincingly shown that anaesthetics block impulses propagated over themedial pathways (RAS), but the laterally conducted impulses remain unim­paired and reach the cortex (French etalI953). This selective susceptibilityof RAS to anaesthctics is due to the path through this being multisynaptic(see above. It is only with larger doses that some depressant action overthalamic relay nuclei is observed.

Feldberg (1959) and his coUeagues (1953) have shown that sleep-like can.ditions can be produced by injecting small doses of drugs into the II I ventricleand aqu.:;duct (Fig. 8.) Such effects can be produced with adrenaline;noradrenaline, Ca cl2, as well as with anaesthetics like Mg CI~, urethane,chloral, and chloralose. The amounts injected are far too small to producesuch effects on intraven()Us injection. These effects must therefore be pro­

~~~:~u~~e('~At~7.cffect on ~~uctures lying just outside the III ventricle and

I 6 CO 'CEPTS OF CONSCIOU N S

I TR TRIC LAR C LR BBER D1APHRAOM. STILETTE & C.-\P.

Fi . B. ia ram illuscral s the method of injeclion into the lateral ,-entricle. Cannula iscrewed into Lhe skull. (From Feldberg and herwood, 1953).

In the end it mu t be admitt d that inspite ofall thi exten ive work, thevariou aspects about the mechanism of consciousnes are by no mean com­pletely understood. Much has been left unsaid and many questions leftunan wered. Only further developments will tell us whether th conceptdeveloped in this review will stand the test of future work.

REFERE 'CE •

1. lien, \\ .. (1932); J. 1V000h. Acad. c., 22,490.2. Anand, B. K. (1955); Ind. J. Mtd. Ru., ~3, 195.

3. Anand, B. K. and Dua, • (1956): Ibid, «, 107.

4. rcluiDi,., and Moruzzi, G. (1953): EEG elin. tUTophyJiof., 5, 235.

5. Berger, H. (1929); Arch. PS;l',hiat,,87, 527.

B, K. ANAND 167

6. Brain, R. (1958): B1IIi", 81, 426.

7. 8u'mer, .'. (1935): C. R. Soc. BioJ. (Paris), 118, Il35.8. Idem. (l936): Wi, 172,460,

9. Idem. (J937): Bull. Arad. ra)'. med. Bt/giqul, 2, 68.

10. C.irns,H. W.(J952): Brai", n, 109.

J I. Cobb, S. (1948): Fonfh/ians tlj Neurtlp')'rhialT)', 4th ed., Williams and Wilkins, Ilahimort'.

12. Collins, W. f., and O'Leary,J. L. (19.14): EliGClin.Nlllrtlph)"JioJ,li,619.

13. Feldberg, W. (1959): Brit. Med. J., Vol. ii, 771.

)'1. ."eldberg, W., and Sherwood, S. L. (1953): J. Phpinf., 120,3.

15. French,J. D. (1952): A.M.A. Arch. Nturllf. & Psyrhial., 68, 727.

1G. French, J. D. (1960): Ua"dblltlk tlj Phpitlltlg)', Sutitl" I, Volume I I, American Phrsio1o·gical Socie,y, Washington.

17. French,J. D., and Magoun, H. W. (1952): Arch. Neurol. & Ps,.rrhiM. Cliicagn, 68,591.

18. French,J. D., Verzeano, M., and Magoun, H, W. (195:1): A.M.A. Arch. Nturol. PSJ·

rhial., 69, 505.

19. french,). D., Von Amerongcn, F. K., and Magoun, H. W. (1952): Ibid, 68, ':177.

20. Gohz, F. (ISn): Pjlugers Arcll.llS. Pilpilll., 51, 570.

21. Hess, W. R. (19#): Utif'. physitlf.Acla., 2, 305.

22. James, W. (1892): Trxl·!JooK tlj Ps)dtllllg)'. Macmillan, London.

21. Jefferson, G. (1938): Arch. Nturol. P.')"fhiat., Chirogo, 40, 857.

24. Idem. (19:'>8): RditufM Fllrmofill,ujflu Brai", Churchill, London.

25. Jefferson, G., and Johnson, R. T. (1950): Folia pSJrhial. ntll'lIl. 53, 306.

2G. Lindsley, D. n., Schreiner, L. 1-1. 1-1., Knowl('5, W. 8., :lnd Magoun, H. W. (1950): EEG

Clin. NrnrllJ"'piol., 2, 483.

27. Livingston, W. K., Haugen, F. P. and Brookhart.j. M. (1954): N""III'1I,), 4, 485.

28. Mago"n, H. W. (1954): Broill Muhanisms tvld CAnscwlISntu. Blackwell, Oxford.

29. Magann, H. W. (1958): Th IVllki>lj'B,ain, ChariesC. Thomas, Springfield.

30. l\foruui, G., and l\Iagoun, H. W. (19"9): EEG Cli".Nturophpi'lI., I, 455.

31. Penfield, W. (19':12): P./Icr"sojOrgallis.citl"i"ch,Ntrrolls Splnn. R('5. pub!. As.!. nerv.

menl. Dis., 30, 513.

32. Rheinberger, M., and Jasper, H. H. (1931): Am". J. PhJsiDl., 119, 186.

33. Schafer, E. A. (1900): TelC/.!JlJ(}klljPllysiology.I'entland, Edinburgh.

34. Segundo,j. P., Arana.iniquez, R., and French,j. D. (1955): J. J{tur/IJI,rg., 1'2,601.